Abstract
In this essay, I argue in favor of a novel interpretation of the semantic theory that can be found in the Later Mohist writings. Recent interpretations by Chad Hansen and Chris Fraser cast the Later Mohist theory as a realist theory; this includes attributing to the Later Mohists what we can call “kind-realism,” the idea that there is some correct scheme of kind-terms that carves the world at its joints. While I agree with Hansen and Fraser that the Later Mohist theory is realist in various ways, I offer challenges to their kind-realist interpretations, and argue instead for a kind-conventionalist interpretation on which there is no fixed, correct set of kinds, leaving schemes of kind terms to be determined by conventional decisions that occur during disputation.